LibreImageDream

Use restrictions apply. These weights descend from Stable Diffusion 2.1 and are distributed under the CreativeML OpenRAIL++-M License (see LICENSE). Its paragraph 5 and Attachment A forbid some uses (among them unlawful content, harming minors, defamation, harassment, discrimination and medical advice); by using the model you agree to them, and you must pass them on to anyone you distribute it to.

ImageDream ("ImageDream: Image-Prompt Multi-view Diffusion for 3D Generation", Peng Wang and Yichun Shi) in the diffusers layout prepared by Jiaxiang Tang for LGM, repackaged for LibreYOLO. From one picture of an object it samples four views around it, which LibreYOLO's LGM family (LibreYOLO/LibreLGM0.4b) turns into 3D Gaussian splats.

from libreyolo import LibreGen

model = LibreGen("lgm")          # downloads this repo and LibreLGM0.4b
model("chair.jpg").save("chair.ply")

This is a mirror, not a new model

Nothing was retrained, converted or re-encoded; every file is byte-identical to upstream and keeps upstream's layout. Upstream's remote-code files (pipeline_mvdream.py, unet/mv_unet.py) and its README are not mirrored: LibreYOLO ships its own implementation.

• Upstream repository: ashawkey/imagedream-ipmv-diffusers
• Upstream revision: 73a034178e748421506492e91790cc62d6aefef5
• Original weights: Peng-Wang/ImageDream (sd-v2.1-base-4view-ipmv.pt), fine-tuned from MVDream and Stable Diffusion 2.1 base
• Paper: arXiv:2312.02201

Folder	Component
unet/	ImageDream multi-view UNet with the image-prompt branch (1.3B parameters with its resampler)
vae/	Stable Diffusion 2.1 VAE
text_encoder/, tokenizer/	OpenCLIP ViT-H/14 text encoder of Stable Diffusion 2.1
image_encoder/, feature_extractor/	LAION CLIP-ViT-H-14 image encoder (laion2B-s32B-b79K, MIT)
scheduler/	DDIM configuration

License

Upstream tags these weights openrail and ships no license text. They are a fine-tune of Stable Diffusion 2.1 base, whose CreativeML OpenRAIL++-M License requires every derivative to carry its use-based restrictions, so this mirror ships that license verbatim (Stability AI, dated November 24, 2022). The image encoder is LAION's MIT-licensed CLIP-ViT-H-14. See NOTICE for attribution and SHA-256 of every file. This is LibreYOLO's reading of the terms, not a statement by the upstream authors.